Participant Preferences for Pharmacologic Chronic Pain Treatment Trial Characteristics: An ACTTION Adaptive Choice-Based Conjoint Study

Accepted Manuscript Participant preferences for pharmacologic chronic pain treatment trial characteristics: an ACTTION adaptive choice-based conjoint study Shannon M. Smith, Jennifer S. Gewandter, Rachel A. Kitt, John D. Markman, Janet A. Vaughan, Penney Cowan, Ernest A. Kopecky, Richard Malamut, Alesia Sadosky, Leslie Tive, Dennis C. Turk, Robert H. Dworkin PII:

S1526-5900(16)30182-1

DOI:

10.1016/j.jpain.2016.07.008

Reference:

YJPAI 3282

To appear in:

Journal of Pain

Received Date: 12 April 2016 Revised Date:

20 June 2016

Accepted Date: 26 July 2016

Please cite this article as: Smith SM, Gewandter JS, Kitt RA, Markman JD, Vaughan JA, Cowan P, Kopecky EA, Malamut R, Sadosky A, Tive L, Turk DC, Dworkin RH, Participant preferences for pharmacologic chronic pain treatment trial characteristics: an ACTTION adaptive choice-based conjoint study, Journal of Pain (2016), doi: 10.1016/j.jpain.2016.07.008. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Page 1 of 21 Participant preferences for pharmacologic chronic pain treatment trial characteristics: an ACTTION adaptive choice-based conjoint study Shannon M. Smitha*†, Jennifer S. Gewandtera*, Rachel A. Kitta, John D. Markmanb, Janet A. Vaughana, Penney Cowanc, Ernest A. Kopeckyd, Richard Malamute, Alesia Sadoskyf, Leslie

a

RI PT

Tivef, Dennis C. Turkg, Robert H. Dworkinh

Department of Anesthesiology, University of Rochester School of Medicine and Dentistry,

Rochester, NY, USA

Department of Neurosurgery, University of Rochester, Rochester, NY, USA

c

American Chronic Pain Association, Rocklin, CA, USA

d

Collegium Pharmaceutical, Inc., Canton, MA, USA

e

Teva Pharmaceuticals, North Wales, PA, USA

f

Pfizer Inc, New York, NY, USA

g

Department of Anesthesiology and Pain Medicine, University of Washington, Seattle, WA,

M AN U

USA h

SC

b

Departments of Anesthesiology and Neurology and Center for Human Experimental

Therapeutics, University of Rochester School of Medicine and Dentistry, Rochester, NY, USA

TE D

* Contributed equally to this work

___________________________

† Corresponding author. Tel.: +1 585 273 2382; fax: +1 585 244 7271

AC C

EP

E-mail address: [email protected] (S.M. Smith)

Number of pages: 21

Number of tables: 4, with 1 Supplementary Table Number of figures: 1, with 1 Supplementary Figure

Running title: Participant preferences for pain trial characteristics

ACCEPTED MANUSCRIPT Page 2 of 21 Abstract Barriers to clinical trial recruitment can delay study completion, potentially resulting in increased costs and an unrepresentative sample. In the current study of 150 participants with chronic pain,

RI PT

we used a computerized adaptive choice-based conjoint (ACBC) survey that included 8 characteristics that may affect enrollment in pharmacologic pain treatment trials (i.e., treatment allocation; frequency of pain ratings; treatment administration method; current medications;

SC

number of study visits; availability of evening and weekend visits; invasiveness of laboratory procedures; payment). These data were analyzed using Sawtooth Software (Orem, UT), which

M AN U

identifies the characteristics that dominate participants’ decisions across multiple sets of potential trials. Three characteristics had the largest relative importance in participants’ trial preferences: (1) invasiveness of required laboratory procedures (i.e., 22%), with no procedures or blood tests preferred over ice water sensory testing or skin biopsy; (2) ability to continue current pain medications (21%); and (3) payment for study participation (21%), with higher

TE D

payment preferred. The fourth most important characteristic was number of study visits (13%), with participants preferring fewer in-person visits and more phone contacts. Understanding the

treatment trials.

AC C

Perspective

EP

preferences of potential participants is an important step toward enhancing enrollment in pain

This article presents the preferences of individuals with chronic pain conditions regarding modifiable pain treatment trial characteristics (e.g., number of study visits; payment; treatment allocation). These findings may help to improve enrollment into analgesic clinical trials and in turn accelerate the development of new pain treatments.

Keywords Pain, randomized controlled trial, recruitment, conjoint analysis

ACCEPTED MANUSCRIPT Page 3 of 21 Introduction Chronic pain is a common and often debilitating condition. It is estimated to affect between 20% and 35% of adults worldwide [4,9,12,15]. Unfortunately, many current treatments

RI PT

for chronic pain are only modestly effective and can cause intolerable adverse effects [29]. Improving the design and execution of randomized clinical trials (RCTs) may help accelerate the discovery of more effective treatments with fewer side effects. Participant recruitment is one of

SC

the largest impediments to RCT completion [16]. A recent study reporting on data extracted from www.clinicaltrials.gov found that 5.6% of the registered analgesic RCTs terminated before

M AN U

the planned number of subjects was accrued, and of those, 38% terminated early due to enrollment difficulties [5]. Without an attempt to understand the trial design features that participants would prefer or dislike [16], enrollment into pain clinical trials will likely continue to be challenging.

Previous research in various therapeutic areas including pain has evaluated patient

TE D

attitudes that affect clinical trial recruitment (e.g., [1,19,25,30,31]). Other studies have investigated the relative value of distinct recruitment strategies and study designs on trial participation (e.g., [2,3,13,14,21,23,28]). Research has also examined predictors of recruitment

EP

in pain RCTs (e.g., degree of pain interference with daily life, sex) [7,27,35]. However, few studies have focused on clinical trial characteristics that investigators can modify and adapt to

AC C

optimize recruitment (see [17] for an exception), and none of these have focused on pain RCTs. Furthermore, no research of which the authors are aware has investigated the effects of clinical trial design characteristics on participant preferences for chronic pain clinical trials. We used conjoint analysis to examine preferences among individuals with chronic pain

regarding clinical trial characteristics that are relevant to chronic pain treatment trials, likely to affect participants’ willingness to enroll in a trial, and are modifiable by study investigators. Conjoint analysis was first developed for marketing research to predict consumer buying preferences [22]. Choice-based conjoint analysis is a specific method that asks consumers to

ACCEPTED MANUSCRIPT Page 4 of 21 choose between hypothetical products (e.g., TVs) comprised of a number of product characteristics (e.g., size; price) and various options within those characteristics (e.g., 42”, 55”, 60”; $300, $500, $700; [22]). Adaptive choice-based conjoint (ACBC) analysis modifies the

RI PT

product characteristics and options that are presented to each survey respondent based on their previous responses, which can increase the depth of data gathered from respondents and lead to better prediction of future behavior [22]. Conjoint analysis has been used previously in

SC

medicine to investigate factors that influence patient decisions regarding medications

[11,18,20,33,34], clinician decisions surrounding treatment options [32], and to explore how trial

M AN U

design might increase enrollment in RCTs of Alzheimer disease [17]. In this study, we used ACBC analysis to identify the preferences of potential participants for various modifiable design characteristics (e.g., amount of payment; number of study visits) of RCTs for chronic pain treatments. We focused on characteristics relevant to pharmacologic pain treatment trials. We limited our survey to phase 3 trials due to larger enrollment requirements than earlier phase

TE D

trials, which makes recruitment challenges particularly problematic.

Methods

EP

This study had two components (i.e., cognitive interviews; ACBC survey) and was approved as an exempt study given minimal risk to participants by the University of Rochester

AC C

Research Subjects Review Board, and therefore did not require a signed informed consent form. Participants received an information letter and had an opportunity to ask any questions before beginning the study. Participation in the cognitive interviews and ACBC survey was not connected with any other research (i.e., participants were informed that they were being asked questions about pain RCTs in general and that we were not recruiting for an RCT).

Cognitive interviews

ACCEPTED MANUSCRIPT Page 5 of 21 In order to develop the characteristics and options that were used to make up hypothetical clinical trials for conjoint analysis, we conducted cognitive interviews with 20 individuals with diverse chronic pain conditions between 07/01/2014 and 09/04/2014.

RI PT

Participants were recruited using a flyer posted at a pain clinic affiliated with the University of Rochester Medical Center (Rochester, NY, USA). Interested individuals called one member of the research team (JSG) who described the cognitive interviews and determined eligibility (i.e.,

SC

experiencing pain for at least 3 months; willingness to consider enrolling in a pharmacologic pain treatment trial without a specific commitment to participate). Eligible and interested

M AN U

individuals were mailed a packet containing a list of 10 characteristics relevant to phase 3 pharmacologic pain treatment trials (e.g., payment) and 2-4 options within each characteristic (e.g., $50 per week plus travel expenses, $25 per week plus travel expenses, only travel expenses). During the cognitive interview, participants were asked what each characteristic meant to them and if they preferred any options within a characteristic. These questions were

TE D

designed to determine whether participants understood what each characteristic and option meant and whether any characteristics were particularly important or unimportant. Next, sets of hypothetical trials were presented in the way that they were to be displayed in the final survey in

EP

order to discover whether any additional instructions might simplify survey completion. Based on participant feedback, the characteristics and options were modified 7 times,

AC C

resulting in a final list of 8 trial characteristics, each with 2-4 options (Table 1). In order to ensure that the included trial characteristics were representative of clinical trial factors that are important to individuals with chronic pain, interviewees were asked whether any additional issues would affect their decision to enroll in a pharmacologic treatment trial for their pain; no further characteristics that were related to modifiable trial design characteristics were suggested. After the final revisions to the characteristics and options, 5 cognitive interviews were conducted and no new themes or questions about the characteristics and options arose. It became apparent in the cognitive interviews that participants did not readily understand that

ACCEPTED MANUSCRIPT Page 6 of 21 they were being asked to choose between hypothetical clinical trials (i.e., the ACBC survey presentation described in the next section). This led to the creation of a script for the conjoint survey in which the researcher demonstrates the process of weighing a hypothetical clinical

Adaptive choice based conjoint (ACBC) survey

RI PT

trial’s characteristics in order to decide whether or not to enroll.

SC

The ACBC survey using Sawtooth Software (Orem, UT) survey methodology was

conducted at 2 pain clinics affiliated with the University of Rochester Medical Center (Rochester,

M AN U

NY) between 12/05/2014 and 09/02/2015. Individuals with chronic pain were recruited at visits to these clinics. We used the following formula [22] to estimate the sample size necessary to detect the main effects of the 8 trial characteristics included in this study: nta/c > 1000, where n=number of participants, t=number of “tournament tasks” (i.e., number of sets of hypothetical clinical trials from which participants chose 1 in order to identify what characteristics participants

TE D

prioritize within the context of an RCT; see Supplementary Figure 1 for a sample tournament task), a=number of hypothetical RCTs per task, and c=highest number of options in any characteristic. This resulted in a minimum sample size estimate of 84 participants, using 16

EP

tournament tasks and 3 alternative RCTs per task, given that the highest number of options per characteristic was 4. In order to have additional power to explore subgroup analyses based on

AC C

participant demographics, our target sample size was 200-250 complete survey responders. Due to recruitment challenges, prior to analyzing any of the data, recruitment was stopped at 150 participants, still well above the minimum sample size requirement for testing the main effects.

Upon check-in at the pain clinic, patients were handed a flyer describing the survey that directed them to inquire further with a research assistant at the pain clinics. Eligible patients (1) were currently experiencing chronic pain and had been for at least 3 months, (2) were at the pain clinic for a visit that did not involve a medication injection (in order to prevent including

ACCEPTED MANUSCRIPT Page 7 of 21 patients who had been sedated), (3) expressed a willingness to consider participating in clinical trials, (4) were able to speak and read English, (5) comfortable using a computer, and (6) anticipated that they had at least 45 minutes available after their visit to complete the survey.

RI PT

Completion of the survey took approximately 45 minutes and participants were paid $20 for their time.

During the ACBC survey, participants were instructed that they should consider all

SC

pharmacologic treatment trials in the survey to be 16 weeks in duration to correspond to phase 3 trials (e.g., 4 weeks for a run-in or titration period, 12 weeks for the double-blind treatment

M AN U

period). Initially, participants were asked a series of questions to determine whether any of the characteristic options were unacceptable or required for their trial participation. Participants then completed the tournament tasks. In each tournament task, participants were presented with 3 hypothetical clinical trials created by combining the various options within each characteristic, and they were asked to select the trial they would most prefer to join (see Supplementary Figure

TE D

1 for an example). Each characteristic option was assigned a utility score, which is a quantification of how frequently it occurred among the clinical trials that the participants selected across all tournament tasks [22]. The broader the range of the utility scores between the options

EP

within a characteristic, the larger effect that characteristic had when respondents chose amongst the hypothetical clinical trials. The ranges of these utility scores were used to calculate

AC C

relative importance scores for each characteristic using the following formula: relative importance of characteristic Y = (The range of utility scores for characteristic Y) / (The sum of the range of utility scores for all of the characteristics) * 100. Therefore, the sum of the relative importance of all of the characteristics used to create the hypothetical clinical trials sums to 100 and can be used to identify which characteristics have the greatest effect on respondents’ preferences regarding clinical trials relative to the other characteristics included in the ACBC survey.

ACCEPTED MANUSCRIPT Page 8 of 21 Exploratory subgroup analyses were conducted to examine differences in the relative importance of the RCT characteristics comparing: (1) males and females, 2) median split age: younger (21 – 48) and older (49 – 75), (3) lower (high school diploma or less) and higher

RI PT

education (college degree or more), (4) average pain intensity over the past week on a 0-10 numerical rating scale (pain intensity <7 vs. > 7), (5) racial background (White/Caucasian vs. other racial backgrounds), (6) chronic pain condition (musculoskeletal vs. neuropathic pain vs.

SC

other), (7) how well pain is controlled (well or completely vs. not at all or slightly), and (8) prior RCT participation (yes vs. no). Due to the large number of comparisons (i.e., 8 RCT

M AN U

characteristics X 8 subgroups), the relevance of inferential statistical tests would be difficult to interpret. Therefore, we solely report descriptive statistics for the subgroup analyses and discuss the largest differences in the Results section.

Using the data collected from the survey, Sawtooth Software was used to simulate participant preferences for hypothetical RCTs in order to predict how variations in trial

TE D

characteristic options could affect recruitment. First, we simulated the “Typical RCT” in which blood tests are the only laboratory tests required, current pain medications (other than acetaminophen or NSAIDs as needed) are prohibited, patients are reimbursed solely for travel

EP

expenses, study appointments are in person every other week, the study drug is taken as a pill twice per day, pain intensity is rated once a day, study visits occur weekdays between 9am and

AC C

5pm, and 50% of participants receive the study drug. We then varied options within the 3 characteristics with the highest relative importance (i.e., nature of required laboratory procedures, ability to continue current pain medications, and payment) to create hypothetical “Alternative RCTs”. The preferences for each Alternative RCT when compared with the Typical RCT are presented in the results.

Results

ACCEPTED MANUSCRIPT Page 9 of 21 In a 9 month period, 312 individuals approached the research assistant to find out more about the study and 150 participants completed the survey (see Figure 1 for participant flow; see Table 2 for demographics).

RI PT

Required and unacceptable characteristic options In the initial stage of the ACBC survey, only a small number of participants indicated that any trial characteristic options would be required or unacceptable when deciding whether to

SC

participate in a clinical trial. The 2 characteristic options that were most frequently listed as required for enrollment into an RCT were being able to take their current pain medications

M AN U

during the study (n=38, 25%) and receiving at least $25 per week plus travel expenses (n=13, 9%). The characteristic options most frequently chosen as unacceptable were: (1) not being able to take current pain medications (n=38, 25%), (2) skin biopsy (n=35, 23%), (3) receiving only travel expenses without weekly reimbursement for time spent on the study (n=13, 9%), (4)

TE D

17 weekly in person study visits (n=8, 5%), and (5) cold water sensory testing (n=7, 5%).

Relative importance of the RCT characteristics

Three characteristics had the highest relative importance, all of which were nearly

EP

identical in importance. The first was the nature of required laboratory procedures (22%; Supplementary Table 1), with participants preferring a study that required no procedures or only

AC C

blood tests over more painful or invasive tests (i.e., cold water sensory testing or skin biopsy, Table 3). The second highest relative importance was the ability to continue using current pain medications (21%), with participants preferring to continue taking these medications. Payment for study participation (21%) had the third highest relative importance. Table 3 shows that participants preferred studies that had higher payment. The timing and type of study visits (13%) was ranked fourth in relative importance, with greater preference for fewer in-person visits and more phone contacts (Table 3). The medication administration method had the fifth highest relative importance (9%), with participants preferring to take a pill twice a day over a pill 3 times

ACCEPTED MANUSCRIPT Page 10 of 21 a day or a cream twice a day (Table 3). Interestingly, the 3 characteristics with the lowest relative importance were how frequently pain is recorded (6%), times available for study visits (5%), and number/percentage of participants receiving the experimental medication (3%).

RI PT

In the exploratory subgroup analyses, the largest numerical difference was seen in the relative importance of the ability to continue taking current pain medications (Supplementary Table 1). For males, the relative importance of this characteristic was 15%, whereas it was 24%

SC

for females. Caucasian participants (24%) and participants with other racial backgrounds (16%) also exhibited different preferences for this characteristic, as did those who had previously

M AN U

participated in RCTs (28%) compared with those who had not (19%), participants with musculoskeletal pain (19%) compared with participants with neuropathic pain (27%), and participants whose pain was well or completely controlled by current treatments (27%) compared with participants whose pain was only slightly or not at all controlled (20%). No substantive numerical differences were seen between participants based on categorical age,

1).

EP

RCT simulations

TE D

education level, average pain intensity over the past week, or pain clinic (Supplementary Table

As the simulations presented in Table 4 show, an Alternative RCT that required skin

AC C

biopsies was much less preferred than the Typical RCT (i.e., 20% who preferred the Alternative RCT vs. 80% who preferred the Typical RCT). Increasing the payment amount to $25 or $50 per week plus travel expenses increased the estimated preference for a trial requiring skin biopsies when compared with the Typical RCT (43% skin biopsies and $25 per week vs. 57% Typical RCT; 54% skin biopsies and $50 per week vs. 46% Typical RCT; Table 4). Similarly, an Alternative RCT with sensory testing using cold water was less desirable (43%) than the Typical RCT (57%). However, the addition of weekly payments increased the desirability of a trial with sensory testing much more than was the case for an RCT with skin biopsies (69% cold water

ACCEPTED MANUSCRIPT Page 11 of 21 sensory testing with $25 per week vs. 31% Typical RCT; 79% cold water sensory testing with $50 per week vs 21% Typical RCT; Table 4). Participant responses indicated that current pain medications were especially important.

RI PT

Compared with the Typical RCT (27%), a trial that allows current pain medications would be preferred (73%; Table 4). Adding weekly payments to an RCT, as well as allowing current pain medications, further increased participant preference compared with a Typical RCT (85% with

SC

$25 per week; 90% with $50 per week; Table 4). In addition, about half of our sample would prefer a trial that only paid travel expenses and allowed current medications (50%) over a trial

M AN U

that paid $50 per week plus travel expenses, but did not allow current medications (50%). Approximately half of the sample (53%; Table 4) would prefer a trial that required skin biopsies, but also allowed current medications, compared with the Typical RCT that did not allow current medications. The addition of weekly payments to a trial that requires skin biopsy and also allows for current medication use would further increase participant preference (66% with $25 per

Discussion

TE D

week; 73% with $50 per week; Table 4).

EP

In this study, we used ACBC methodology to examine preferences for RCT characteristics among individuals with chronic pain conditions who would consider participating

AC C

in pharmacologic pain treatment trials. Compared with the other characteristics included in this survey, those that most affected participants’ decisions regarding enrollment were: (1) laboratory procedures, with the lowest preference for skin biopsy; (2) ability to continue taking current pain medications; and (3) payment, with more payment preferred over less. Implementing these characteristics may not always be feasible in pain RCTs, as is the case when current pain medications cause unwanted drug interactions with the study drug or confound the treatment effect of the study drug unless studying an adjunct treatment. In addition, internal review boards at different institutions regulate participant compensation to

ACCEPTED MANUSCRIPT Page 12 of 21 varying degrees, which may limit study recruitment. Although it is important to ensure that the amount of payment does not present undue, unethical inducement for individuals to participate against their better judgment [6,10], adequate compensation will likely maximize recruitment

RI PT

and can be justified as compensation for participants’ time and effort. The characteristic with the fourth highest impact on participants’ preferences was the frequency and type of study visits, with fewer in-person visits preferred. Minimizing in-person

SC

trial visits has the potential to increase enrollment and may have the added benefit of improving RCT retention [26]. Less frequent contact between study staff and participants may also

M AN U

decrease the placebo effect [8,24]. It is important to consider that efforts to keep participants engaged in RCTs in lieu of in-person study visits may be needed (e.g., phone calls, newsletters, online study visits)[28].

The remaining 4 RCT characteristics (i.e., how the medication is taken; how frequently pain is recorded; times available for study visit; percent getting experimental treatment) were

TE D

relatively less important in determining participants’ RCT choices. It is surprising that participants’ RCT selections were influenced much less by the percent of participants receiving the experimental study treatment relative to payment for study visits, required laboratory

EP

procedures, the ability to continue their current medications, and the frequency and type of study visits. Individuals who participate in pain RCTs are likely motivated, at least in part, to find

AC C

a more effective treatment [31], and therefore the chances of receiving the experimental treatment might have been expected to have a large effect on RCT selection. In exploratory subgroup analyses, females, Caucasian participants, individuals who had

previously participated in RCTs, individuals with neuropathic pain, and individuals whose pain was well or completely controlled had a higher preference for current pain medications than did males, participants from other racial backgrounds, those who had never participated in RCTs, individuals with musculoskeletal pain, and individuals whose pain was slightly or not at all controlled, respectively. Predictably, individuals whose pain was well controlled valued

ACCEPTED MANUSCRIPT Page 13 of 21 continuation of current pain medications more so than individuals whose pain was poorly controlled. The reasons for the remaining differences are not readily apparent. Other demographic features did not have a considerable influence on preferences. Given that multiple

RI PT

comparisons were made, these findings could be due to random noise and need to be investigated further in future studies.

In the simulation analyses, the Alternative RCT that included skin biopsies was selected

SC

much less frequently than the Typical RCT. Including weekly payments led to an increase in the estimated percentage of participants who would select the Alternative RCT involving a skin

M AN U

biopsy, although only 54% of participants were expected to prefer an Alternative RCT with skin biopsies and $50 per week payments compared with the typical RCT. Cold water sensory testing was less of a disincentive. Although less than half of the participants were expected to choose a trial that required cold water sensory testing over the Typical RCT, the addition of $50 weekly payments to the trial with cold water testing increased this percent to nearly 80%.

TE D

Researchers should consider implementing the least aversive laboratory testing that can answer the question of interest. When more invasive procedures or aversive assessments are necessary, our data suggest that increasing the payment that participants receive may

EP

considerably increase trial recruitment, keeping in mind that payment should not constitute an undue, unethical influence [6].

AC C

Compared with the Typical RCT, nearly three-quarters of the participants were estimated to prefer a trial allowing current pain medications, with $50 weekly payments increasing these estimates to 90%. Interestingly, allowing current pain medications in an Alternative RCT that also required skin biopsies led to estimated preferences (53%) that were similar to preferences for a trial with skin biopsies and $50 weekly payments (54%). Whenever possible, allowing participants to stay on stable dosages of their current pain medications may help in recruiting for chronic pain treatment trials, as well as increasing the generalizability of the trial results.

ACCEPTED MANUSCRIPT Page 14 of 21 However, it is important to weigh these benefits with the potential for current medications to decrease assay sensitivity [8]. This study was novel in examining participant preferences for pain treatment trial

RI PT

characteristics using an ACBC survey design, in which participants were asked to choose whether they would participate in hypothetical RCTs. This method can evaluate the relative importance of each characteristic to participants, as well as how combinations of characteristics

SC

might affect participants. ACBC analysis may more accurately identify participant preferences than would be the case if participants were simply asked to rate the importance of each isolated

M AN U

characteristic, in which case each characteristic could be rated as equally important without any discrimination among the characteristics [22]. Given recruitment challenges [16], implementing the preferred characteristics identified in this study whenever possible may promote RCT enrollment among individuals with chronic pain. In addition, when participant preferences are prioritized to enhance enrollment, it is likely to improve study retention as well.

TE D

It is important to acknowledge this study’s limitations as well. We only recruited at 2 pain clinics in 1 western NY city. This could lead to differences between our study sample and the general population of individuals with chronic pain in ethnicity, culture, and economic

EP

distribution, as well as severity of pain (e.g., pain may be more severe among pain clinic patients, leading to a higher preference for current pain medications). In addition, fewer than

AC C

half of the individuals screened for eligibility participated in this survey which could have resulted in selection bias. For example, recruiting for this study occurred during weekday working hours. This may have resulted in a sample that includes a higher proportion of individuals who are unemployed, out of the work force, or whose work schedules could accommodate RCT participation during weekday working hours, and as such, the times available for study visits characteristic may be less relevant to our sample than to the population of individuals with chronic pain. Furthermore, a minimum of 45 minutes was required for participation after the clinic appointment, and therefore study participants may differ from

ACCEPTED MANUSCRIPT Page 15 of 21 individuals who were otherwise eligible but did not meet this requirement. Survey participants were paid, possibly inflating the relative importance of the trial payment characteristic in our survey by selecting participants who value payment. Alternatively, if the $20 payment was

RI PT

insufficient compensation for individuals who highly value payment, our results may underestimate the importance of trial payment. It is also possible that study participants were unfamiliar with certain characteristics or options (e.g., biopsy of deep skin layers with needle), or

SC

may have struggled to comprehend the survey. However, the wording was optimized for

understanding based on feedback from chronic pain participants during the cognitive interviews.

M AN U

It is important to bear in mind that the preferences for RCT characteristics are relative to the specific characteristics and options we selected for this survey. We attempted to identify the characteristics and options that were most relevant to typical phase 3 pharmacologic pain treatment trials; however, if other characteristics and options were included, participant preferences may have varied. We planned to enroll a larger number of participants than needed

TE D

to adequately test main effects in order to have sufficient data to observe differences between subgroups. We were unable to reach our initial recruitment goal and therefore had less power to detect differences between subgroups. Finally, the results of this study are exploratory and are

EP

based on survey responses to hypothetical chronic pain treatment trials rather than participant

AC C

enrollment in actual clinical trials.

Conclusions

RCT characteristics that promote assay sensitivity can be inconsistent with the

preferences of participants. For example, allowing current medications could decrease the ability to demonstrate a treatment effect [8], but increase the likelihood of enrollment of a patient into an RCT. Although researchers design their trials in ways to optimize assay sensitivity and internal validity, it is important to balance participant preferences with methodological and regulatory priorities in order to complete trials. The current study provides insights into the

ACCEPTED MANUSCRIPT Page 16 of 21 preferences of potential participants in pain treatment trials. Understanding these preferences may allow investigators to choose clinical trial methods that will maintain assay sensitivity but also boost enrollment, potentially increasing the speed with which new analgesic medications

AC C

EP

TE D

M AN U

SC

RI PT

can be evaluated.

ACCEPTED MANUSCRIPT Page 17 of 21 Disclosures The views expressed in this article are those of the authors, none of whom have financial conflicts of interest specifically related to the issues discussed in this article. Financial

RI PT

support for this project was provided by the Analgesic, Anesthetic, and Addiction Clinical Trial Translations, Innovations, Opportunities, and Networks (ACTTION) public-private partnership with the US Food and Drug Administration (FDA). ACTTION has received research contracts,

SC

grants, or other revenue from the FDA, multiple pharmaceutical and device companies, and other sources. No official endorsement by the FDA or the pharmaceutical and device companies

M AN U

that have provided unrestricted grants to support the activities of ACTTION should be inferred.

Acknowledgements

The preparation of this article was undertaken by ACTTION’s Community Patient Awareness About Clinical Trials (COMPAACT) Working Group and the manuscript was

TE D

reviewed and approved by ACTTION’s Executive Committee. We thank James Koch, BS, from

AC C

EP

the University of Rochester for his assistance with formatting the references.

ACCEPTED MANUSCRIPT Page 18 of 21

AC C

EP

TE D

M AN U

SC

RI PT

References 1. Avis NE, Smith KW, Link CL, Hortobagyi GN, Rivera E. Factors associated with participation in breast cancer treatment clinical trials. J Clin Oncol 24:1860-7, 2006. 2. Bell-Syer SEM, Moffett JAK. Recruiting patients to randomized trials in primary care: principles and case study. Family Practice 17:187-191, 2000. 3. Bhanushali MJ, Gustafson T, Powell S, Conwit RA, Wolinsky JS, Cutter GR, Lublin FD, Cofield SS. Recruitment of participants to a multiple sclerosis trial: the CombiRx experience. Clin Trials 11:159-66, 2014. 4. Breivik H, Eisenberg E, O'Brien T, Openminds. The individual and societal burden of chronic pain in Europe: the case for strategic prioritisation and action to improve knowledge and availability of appropriate care. BMC Public Health 13:1229, 2013. 5. Cepeda MS, Lobanov V, Berlin JA. Use of ClinicalTrials.gov to estimate conditionspecific nocebo effects and other factors affecting outcomes of analgesic trials. J Pain 14:405-11, 2013. 6. Council for International Organizations of Medical Sciences. International ethical guidelines for biomedical research involving human subjects, 2002. Available from: http://www.cioms.ch/publications/guidelines/guidelines_nov_2002_blurb.htm. Accessed 2/3/2016. 7. Creel AH, Losina E, Mandl LA, Marx RJ, Mahomed NN, Martin SD, Martin TL, Millett PJ, Fossel AH, Katz JN. An assessment of willingness to participate in a randomized trial of arthroscopic knee surgery in patients with osteoarthritis. Contemp Clin Trials 26:169-78, 2005. 8. Dworkin RH, Turk DC, Peirce-Sandner S, Burke LB, Farrar JT, Gilron I, Jensen MP, Katz NP, Raja SN, Rappaport BA, Rowbotham MC, Backonja MM, Baron R, Bellamy N, Bhagwagar Z, Costello A, Cowan P, Fang WC, Hertz S, Jay GW, Junor R, Kerns RD, Kerwin R, Kopecky EA, Lissin D, Malamut R, Markman JD, McDermott MP, Munera C, Porter L, Rauschkolb C, Rice AS, Sampaio C, Skljarevski V, Sommerville K, Stacey BR, Steigerwald I, Tobias J, Trentacosti AM, Wasan AD, Wells GA, Williams J, Witter J, Ziegler D. Considerations for improving assay sensitivity in chronic pain clinical trials: IMMPACT recommendations. Pain 153:1148-58, 2012. 9. Elzahaf RA, Tashani OA, Unsworth BA, Johnson MI. The prevalence of chronic pain with an analysis of countries with a Human Development Index less than 0.9: a systematic review without meta-analysis. Curr Med Res Opin 28:1221-9, 2012. 10. Emanuel EJ. Ending concerns about undue inducement. J Law Med Ethics 32:100-5, 2004. 11. Gelhorn HL, Stringer SM, Brooks A, Thompson C, Monz BU, Boye KS, Hach T, Lund SS, Palencia R. Preferences for medication attributes among patients with type 2 diabetes mellitus in the UK. Diabetes Obes Metab 15:802-9, 2013. 12. Goldberg DS, McGee SJ. Pain as a global public health priority. BMC Public Health 11:770, 2011. 13. Harrison T, Miyahara S, Lee A, Evans S, Bastow B, Simpson D, Gilron I, Dworkin R, Daar ES, Wieclaw L, Clifford DB, Team AA. Experience and challenges presented by a multicenter crossover study of combination analgesic therapy for the treatment of painful HIV-associated polyneuropathies. Pain Med 14:1039-47, 2013. 14. Holman AJ, Neradilek MB, Dryland DD, Neiman RA, Brown PB, Ettlinger RE. Patientderived determinants for participation in placebo-controlled clinical trials for fibromyalgia. Curr Pain Headache Rep 14:470-6, 2010. 15. Institute of Medicine. Relieving pain in America: a blueprint for transforming prevention, care, education, and research. Washington, DC: The National Academies Press, 2011.

ACCEPTED MANUSCRIPT Page 19 of 21

22. 23.

24.

25. 26.

27.

28.

29. 30.

31. 32. 33.

34.

RI PT

21.

SC

20.

M AN U

19.

TE D

18.

EP

17.

Institute of Medicine. Envisioning a Transformed Clinical Trials Enterprise in the United States: Establishing an Agenda for 2020: Workshop Summary. Washington, DC: The National Academies Press, 2012. Karlawish J, Cary MS, Rubright J, Tenhave T. How redesigning AD clinical trials might increase study partners' willingness to participate. Neurology 71:1883-8, 2008. Kleinman L, McIntosh E, Ryan M, Schmier J, Crawley J, Locke GR, 3rd, De Lissovoy G. Willingness to pay for complete symptom relief of gastroesophageal reflux disease. Arch Intern Med 162:1361-6, 2002. Maida S, Dalla Costa G, Rodegher M, Falautano M, Comi G, Martinelli V. Overcoming recruitment challenges in patients with multiple sclerosis: Results from an Italian survey. Clin Trials 11:667-72, 2014. Marti J. Assessing preferences for improved smoking cessation medications: a discrete choice experiment. Eur J Health Econ 13:533-48, 2012. Myles PS, Fletcher HE, Cairo S, Madder H, McRae R, Cooper J, Devonshire D, Hunt JO, Richardson J, Machlin H, Morgan EB, Moloney J, Downey G. Randomized trial of informed consent and recruitment for clinical trials in the immediate preoperative period. Anesthesiology 91:969-78, 1999. Orme BK. Getting started with conjoint analysis: strategies for product design and pricing research. Manhattan Beach, CA: Research Publishers LLC, 2014. Peters-Lawrence MH, Bell MC, Hsu LL, Osunkwo I, Seaman P, Blackwood M, Guillaume E, Bellevue R, Krishnamurti L, Smith WR, Dampier CD, Minniti CP, Sickle Cell Disease Clinical Research N. Clinical trial implementation and recruitment: lessons learned from the early closure of a randomized clinical trial. Contemp Clin Trials 33:291-7, 2012. Posternak MA, Zimmerman M. Therapeutic effect of follow-up assessments on antidepressant and placebo response rates in antidepressant efficacy trials: metaanalysis. Br J Psychiatry 190:287-92, 2007. Ransom S, Azzarello LM, McMillan SC. Methodological issues in the recruitment of cancer pain patients and their caregivers. Res Nurs Health 29:190-8, 2006. Rutherford BR, Cooper TM, Persaud A, Brown PJ, Sneed JR, Roose SP. Less is more in antidepressant clinical trials: a meta-analysis of the effect of visit frequency on treatment response and dropout. J Clin Psychiatry 74:703-15, 2013. Salomons TV, Wowk AA, Fanning A, Chan VW, Katz J. Factors associated with refusal to enter a clinical trial: epidural anesthesia is a deterrent to participation. Can J Anaesth 49:583-7, 2002. Tsai PF, Chang JY, Chowdhury N, Beck C, Roberson PK, Rosengren K. Enrolling older adults with cognitive impairment in research: lessons from a study of Tai Chi for osteoarthritis knee pain. Res Gerontol Nurs 2:228-34, 2009. Turk DC, Wilson HD, Cahana A. Treatment of chronic non-cancer pain. Lancet 377:2226-35, 2011. van den Berg L, Lobatto RM, Zuurmond WW, de Lange JJ, Wagemans MF, Bezemer PD. Patients' refusal to participate in clinical research. Eur J Anaesthesiol 14:287-9, 1997. Wasan AD, Taubenberger SP, Robinson WM. Reasons for participation in pain research: can they indicate a lack of informed consent? Pain Med 10:111-9, 2009. Whitman CB, Shreay S, Gitlin M, van Oijen MG, Spiegel BM. Clinical factors and the decision to transfuse chronic dialysis patients. Clin J Am Soc Nephrol 8:1942-51, 2013. Wittink MN, Cary M, Tenhave T, Baron J, Gallo JJ. Towards Patient-Centered Care for Depression: Conjoint Methods to Tailor Treatment Based on Preferences. Patient 3:145157, 2010. Wittink MN, Morales KH, Cary M, Gallo JJ, Bartels SJ. Towards personalizing treatment for depression : developing treatment values markers. Patient 6:35-43, 2013.

AC C

16.

ACCEPTED MANUSCRIPT Page 20 of 21

EP

TE D

M AN U

SC

RI PT

Yuasa H, Kurita K, Westesson PL. External validity of a randomised clinical trial of temporomandibular disorders: analysis of the patients who refused to participate in research. Br J Oral Maxillofac Surg 41:129-31, 2003.

AC C

35.

ACCEPTED MANUSCRIPT Page 21 of 21

AC C

EP

TE D

M AN U

SC

RI PT

Fig 1. Participant flow

ACCEPTED MANUSCRIPT

Table 1 Study characteristics and options as presented to patients OPTION 1

OPTION 2

OPTION 3

How you take the medicine

Cream 2 times per day

Pill 2 times per day

Pill 3 times per day

Number (%) of participants who get experimental drug (as opposed to placebo drug)

1 out of 2 (50%)

2 out of 3 (67%)

Timing and type of study visits (over 16 weeks)

Once a week appointments (17 in person appointments)

Every other week appointments (6 phone appointments and 3 in person appointments)

How often you have to record your pain (0-10) for the study

1 time per week

Times available for study visits

Weekdays between 9am and 5pm

SC Every other week appointments (9 in person appointments)

M AN U 1 time per day

TE D

EP None

AC C

Laboratory procedures at the first and last study visits

OPTION 4

RI PT

CHARACTERISTIC

3 times per day

Weekdays between 9am and 8pm

Weekdays between 9am and 8pm or weekends

Blood test

Test to determine how long you can keep your hand in ice water

Whether you can take your current pain medications during the study

Yes

No, but you can take medications like Tylenol or Advil as needed

Payment

Only travel expenses

$25 per week plus travel expenses

$50 per week plus travel expenses

Biopsy of deep skin layers with needle

ACCEPTED MANUSCRIPT

Table 2 Demographics Age in years 47.7 (12.2)

Range

21 – 75

Pain intensity rating (0-10 NRS) Mean (SD)

RI PT

Mean (SD)

7.0 (2.0) 2 – 10

Length of time experiencing pain condition in years Median (Interquartile range)

SC

Range

7.0 (3.5-13.2)

0.25 – 47

M AN U

Range

Sex Female Male

TE D

Race

N (% out of 150)

108 (72%) 42 (28%)

100 (67%)

Black or African American

31 (21%)

Hispanic

8 (5%)

Multiple Asian

EP

White or Caucasian

AC C

Native American

5 (3%) 1 (1%) 1 (1%)

Chronic pain condition Low back pain

63 (42%)

Musculoskeletal (not specified as low back or osteoarthritis pain)

26 (17%)

Neuropathic pain

20 (13%)

Other

13 (9%)

Osteoarthritis

9 (6%)

ACCEPTED MANUSCRIPT

Fibromyalgia

7 (5%)

Multiple

7 (5%)

Chronic regional pain syndrome

5 (3%)

RI PT

How well pain is controlled by current treatments Completely controlled

2 (1%)

Well controlled

29 (19%)

Slightly controlled

95 (63%) 18 (12%)

SC

Not controlled at all No current pain treatments

6 (4%)

1 2 3 4 Other (no number specified)

TE D

No prior RCT participation

M AN U

Number of prior randomized clinical trials participants enrolled in

19 (13%) 10 (7%) 2 (1%) 1 (1%) 2 (1%) 116 (77%)

AC C

EP

SD=standard deviation; NRS=numerical rating scale, RCT=randomized controlled trial.

ACCEPTED MANUSCRIPT

Table 3 Average utilities of options within trial characteristics Average utility

RI PT

Laboratory procedures at the first and last study visits None

143

Blood test

139

M AN U

Biopsy of deep skin layers with needle

SC

Test to determine how long you can keep your hand in ice water

112

6

Whether you can take your current pain medications during the study Yes No

TE D

Payment

180 20

$50 per week plus travel expenses

176

$25 per week plus travel expenses

104

EP

Only travel expenses

19

AC C

Timing and type of study visit

Every other week appointments (6 phone appointments and 3 in person appointments; 9 total)

138

Every other week appointments (9 in person appointments)

106

Once a week appointments (17 in person appointments) How you take the medicine

56

ACCEPTED MANUSCRIPT

110

Pill 3 times per day

100

Cream 2 times per day

RI PT

Pill 2 times per day

90

How often you have to record your pain (0-10) for the study

110

SC

1 time per week

3 times per day Times available for study visits

M AN U

1 time per day

107 83

105

Weekdays between 9am and 8pm

101

TE D

Weekdays between 9am and 8pm or weekends

Weekdays between 9am and 5pm

94

EP

Number (%) of participants who get experimental drug (as opposed to placebo drug)

AC C

2 out of 3 (67%)

1 out of 2 (50%)

107

93

Utility score=a quantification of how frequently each option occurred among the clinical trials that the participants selected across all tournament tasks [23]; SD = standard deviation. Note: A constant of 100 was added to each utility. The option with the lowest utility within a characteristic was not necessarily disliked by survey participants. Rather, it is the option that was least preferred relative to the other options within the characteristic.

ACCEPTED MANUSCRIPT

Typical RCT:

The medication is a pill 2 times per day

Pain is rated 1 time per day

M AN U

SC

Every other week appointments (9 in person appointments)

Only travel expenses are provided

Alternative RCT Typical RCT with skin biopsy

EP

Typical RCT with skin biopsy and $25 per week

TE D

Blood tests at first and last study visits

Participants cannot take their pain medications, but can take medications like Tylenol or Advil as needed

RI PT

Table 4 Results of simulation studies comparing Alternative RCT to Typical RCT

AC C

Typical RCT with skin biopsy and $50 per week

Weekdays between 9am and 5pm

1 out of 2 (50%) participants get the study drug

Estimated percent of participants who prefer the Alternative RCT over the Typical RCT† 20% 43% 54%

Typical RCT with cold water sensory testing

43%

Typical RCT with cold water sensory testing and $25 per week

69%

Typical RCT with cold water sensory testing and $50 per week

79%

ACCEPTED MANUSCRIPT

73%

Typical RCT with current medications and $25 per week

85%

RI PT

Typical RCT with current pain medications

Typical RCT with current medications and $50 per week

90%

SC

Typical RCT with skin biopsy and current pain medications

M AN U

Typical RCT with skin biopsy, current pain medications, and $25 per week Typical RCT with skin biopsy, current pain medications, and $50 per week †

53% 66% 73%

AC C

EP

TE D

The percent that prefer the Typical RCT can be calculated by subtracting the percent that prefer the Alternative RCT from 100. RCT=randomized controlled trial. Note: The Alternative RCT is the same as the Typical RCT, except the specified trial characteristic options have been changed. For example, “Typical RCT with skin biopsy” is the Typical RCT with a skin biopsy instead of a blood test at the first and last study visits.

ACCEPTED MANUSCRIPT

RI PT

Assessed for eligibility (n=312)

• • •

AC C

EP

Survey completers used in analyses (n=150)

Did not have time to take survey (n=95) Not interested in completing survey (n=5) Could not participate due to current pain (n=2) Started, but did not complete survey (n=12)

TE D

•

M AN U

SC

Excluded (n=162) • Ineligible (n=48) o Unwilling or unable to participate in RCTs (n=17) o Sedating procedure (n=15) o Difficulties with computers (n=8) o Previously completed the survey (n=4) o Not pain clinic patient (n=2) o Pain had subsided (n=1) o Pain for less than 3 months (n=1)

ACCEPTED MANUSCRIPT

Highlights More aversive lab tests were deterrent to participants.

•

Greater study payment motivated participants’ RCT selection.

•

Continuing concomitant pain medications was an incentive for participants.

•

Concomitant medications and payment may counteract aversive lab tests.

AC C

EP

TE D

M AN U

SC

RI PT

•